CodonCode Corporation will develop software modules for the automated analysis of heterozygous insertions and deletions (indels) in large-scale resequencing projects. While insertions and deletions constitute a significant portion of the total genetic variation, homozygous variants of indel mutations are often rare and therefore not seen in resequencing projects. This makes it essential to correctly identify heterozygous indels. However, while several useful software tools for the analysis of heterozygous point mutations are available, no satisfactory tools for automatic detection and analysis of heterozygote indels currently exist. CodonCode Corporation has developed several algorithms for the interactive detection and analysis of heterozygous indels in DNA sequence chromatogramms. The goal of this proposal is to adapt these tools for large-scale resequencing projects. Specifically, CodonCode will (a) develop software modules that can serve as plug-in modules in automated processing pipelines, (b) improve the algorithm performance to meet the needs of large-scale projects, and (c) develop quantitative measures for the accuracy and sensitivity of the heterozygous indel detection and sizing algorithms. The software tools that will be developed under this grant will be able to automatically deconvolute mixed DNA sequence chromatograms where the mixing is due to different causes. While the primary goal is to support the automated analysis of small heterozygous indels in genomic DNA, our tools will also enable the identification of major transcript variants in RT-PCR based resequencing studies, including large rearrangements that would be missed in typical exon-based resequencing studies. The software developed under this grant application will enable us to obtain a complete picture of the variation in the human genome. It will provide essential tools required to analyze insertion and deletion mutations, a class of mutations that is biologically and medically relevant in many different disease areas. [unreadable] [unreadable] [unreadable]